Tesla and supercomputers at full speed

WEEKLY UPDATE

VIP visit in a leather jacket

On Tuesday, October 22, Denmark, where NewDeal Invest is based, received a high-profile visit from Jensen Huang, the founder of NVIDIA. Alongside King Frederik and Director of Danish Center for AI, Nadia Carlsten, Huang inaugurated Denmark’s new supercomputer, a joint initiative between the DCAI, the Novo Nordisk Foundation, and the Danish Export and Investment Fund, which pooled the necessary 700 million DKK to fund it.

This cutting-edge computer, named Gefion after the Norse goddess of fertility, is reported to be 10,000 times faster than current technology available in the country. Equipped with 1,528 NVIDIA H100 GPUs, it represents the pinnacle of today’s computational power. Unfortunately, Gefion's precise location is a secret, so for those interested in seeing this marvel in person, the mystery remains.

Gefion's purpose is far-reaching. It will support advancements in disease diagnosis, drug development, natural disaster forecasting, weather prediction, green solutions, and countless other applications where data analysis requires exceptional computational power.

Diverging U.S. and European Economic Trends

U.S. interest rates continued to climb this week, with minimal impact on European rates. This disparity hasn’t significantly affected stock markets, but in the currency market, it has led to the dollar’s highest value against the euro since early August.

A contributing factor has been the recent rally for U.S. presidential candidate Donald Trump. Markets anticipate that a Trump presidency could result in tax cuts, which might stimulate the economy while also potentially rekindling inflation concerns.

The past month has generally brought positive economic data from the U.S. Early October saw a surprisingly high 254,000 new jobs reported, with gains across various sectors, including cyclical ones like real estate. This points to a robust labor market and a strong economy. Further supporting this outlook are inflation numbers suggesting that price declines in the goods sector have leveled off, which may mean that anticipated interest rate cuts are unlikely in the near term.

In contrast, Europe’s economic situation is different, with inflation trending downward and economic indicators remaining weak. For instance, the latest PMI for Europe was 50, hinting at stagnant growth in the Eurozone. This difference in trajectory has reignited discussions of a possible 0.5% interest rate cut by the ECB in its next meeting.

NDI-FUTURETECH

NDI-FutureTech has experienced a 27,7% increase year-to-date. One of our highest weighted positions today is Tesla which has had a big week. The stock has risen more than 20% following its most recent earnings report, thanks to a higher gross margin. For this reason, we will be highlighting this particular company, as Tesla automotive division retains high profits as other automakers struggle. Why is this? CIO Mads Christiansen explains below:

Why Tesla is poised for success

The answer is quite simple: Tesla is much better at producing cars cost-effectively compared to competitors like the German OEMs (original equipment manufacturers). This gives Tesla the flexibility to choose between high margins on their vehicles or lowering their prices. Tesla consistently chooses to lower prices, which confuses many investors. The reason for this confusion is that price cuts are often interpreted as a sign of tough competition or an inability to sell the product at a premium. In other words, price cuts can be seen as a weakness.

But price cuts can also be a deliberate strategy. Tesla lowers its prices to boost volume, making it more affordable for more people to buy a Tesla. Increased sales volume then lowers production costs even further due to the economies of scale in both manufacturing and distribution.

Why is it cheaper to produce a Tesla than other cars?

The answer lies in Tesla’s innovation in car manufacturing. Tesla’s core focus is on advancing the production process, which is what sets it apart from German automakers.

In the traditional auto industry, there are many stages in the value chain from concept to consumer-ready vehicle. I am not an auto industry expert, so bear with me if there are any inaccuracies here; the key is the concept. In the traditional industry, a designer creates a concept, which engineers then develop. The factory then assesses production feasibility, suppliers are contacted, and after several cycles, the car is ready for production. This process can take years, and once production begins, changes are difficult without consulting designers, engineers, suppliers, and production managers. The value chain is modular and bureaucratic.

Economist Clayton Christensen describes a mechanism in which, once a product is “good enough” to meet consumer needs, the industry’s focus shifts from functionality to design and branding. Cars have been “good enough” for some time, with German OEMs focusing on design, branding, and marketing. Meanwhile, Japanese and Korean automakers have made "good enough" cars focused on price, commoditizing the lower end of the market. German automakers moved upmarket, focusing on higher-priced, high-margin vehicles. This upmarket movement is attractive economically, as it reduces working capital requirements by deprioritizing low-margin segments, improving ROIC (return on invested capital). However, over time, this strategy diminishes the ability to produce cost-competitively. This is precisely how American automakers lost to Toyota, Hyundai, and Kia in the 20th century.

Before Tesla, German automakers gradually lost the ability to produce cars cheaply and efficiently because they didn’t need efficiency when moving upmarket, selling more expensive, high-margin vehicles that reduced working capital and improved ROIC.

Built from the ground up

Tesla’s production system was designed from the ground up 10-15 years ago. Elon Musk is a first-principles thinker, and one of his key tenets is that designers and production must operate under one roof so that design integrates smoothly into manufacturing. This eliminates bureaucracy and modularity. Tesla’s cars are designed with cost-effective production in mind.

Iterating and experimenting are crucial to development, and Tesla is set up for this. Musk often mentions finding savings "penny by penny." This means that design and production continuously evolve to deliver incremental savings and improvements, which explains why Tesla’s production costs keep decreasing.

The new trend in production and design is digital twins. Tesla’s vehicles, supply chains, and production are powered by digital twins. This allows for real-time monitoring, immediate detection of faults, and ongoing optimization, as data flows into the twin and informs the system. It enables simulation of solutions or improvements, predicting issues before implementation. These two factors reduce the risk of iterative production changes, making small, ongoing improvements possible.

Swedish company Hexagon, a supplier to BYD, develops digital twins for manufacturing. Josh Weiss, president of Hexagon's Manufacturing Intelligence division, comments:

“It’s ironic that manufacturing invented the automation and agile practices that are driving business transformation in other industries and is now struggling to transform – but that’s because achieving digitalization throughout manufacturing value chains is a very real, complex, and human challenge. Digital twins are crucial to making factories smart, they enable teams to solve problems across departments, to innovate, and that same high-quality data underpins productivity-superchargers such as AI and robotics."

Auto Manufacturing to resemble Microchip Production

When TSMC starts producing a new generation of chips, the gross margin drops initially, then steadily rises until the next generation begins. At first, chips may even sell below production cost. A famous example is Fairchild Semiconductor in the 1950s. Their integrated circuits were so expensive that only the U.S. military and NASA could afford them. Fairchild Semiconductor drastically lowered prices, which opened up the consumer electronics market. With price reductions came volume, and with volume came lower production costs. Tesla follows this same playbook for similar reasons.

In my view, cars resemble microchips in that volume is crucial to competitiveness. Producing in small volumes results in high production costs per unit, but large volumes lower costs per unit due to two factors:

1. Fixed costs spread across more units
2. Production optimization improves efficiency as volume increases

Note that point two doesn’t apply to traditional car production, where incremental improvements are stifled by bureaucracy and lack of digital twins.

In the chip industry, it’s well-known that profits on a chip increase the longer it’s in production. I believe that in Tesla’s era, cars will have similar economic dynamics, making the car industry more like a volume game. We might even see car production follow the path of chip manufacturing, with a few large “foundries” producing cars (like TSMC, Samsung, Intel, Global Foundries, and Super Micro Computer in chips) while many companies design but don’t manufacture. I don’t think global OEMs can maintain competitive in-house production of cars, and I suspect the future for some automakers will be in design-only roles, akin to “fabless” companies in the chip world.

Right now, Tesla is winning. They are cutting prices and selling more cars, which reduces the cost per unit, allowing for further price cuts and more sales. Design, marketing, and branding matter less when Tesla offers pure value to customers seeking transportation. Meanwhile, European OEMs are seeing their sales drop, which raises their production costs, and they can’t keep up with price reductions. Tesla is in a positive spiral driven by price cuts, while German OEMs are in a negative one, fueled by Tesla’s pricing strategy.

THE WEEK AHEAD

A Major Week for Earnings

Next week promises a busy schedule for earnings reports, particularly among key members of the “Magnificent 7” and other large companies. Highlights include Alphabet, reporting Tuesday, followed by Microsoft and Meta on Wednesday, and Amazon and Apple on Thursday. These reports could provide insights into how tech giants are handling an uncertain economic landscape. We’ll also monitor AMD, Mercado Libre, Intel, Onsemi, Carvana, FuboTV, Super Micro Computer, Uber, and Mobileye.

U.S. Economic Data Ahead of the Election

Market attention will also be focused on economic indicators that could influence the Federal Reserve’s stance on inflation. Key releases next week include U.S. manufacturing PMI, initial and continuing jobless claims on Thursday, and the pivotal non-farm payrolls on Friday, November 1. This week also marks the final stretch before the U.S. presidential election on November 5. Expect more insights on our take in next week’s update.